Supply and Demand 2001

Supply and Demand for Software Developers in India
By
Ronald Fernandes
Ashish Arora
H. John Heinz III School of Public Policy & Management, Carnegie Mellon University.
Pittsburgh, PA. 15213.
Jai Asundi
Department of Engineering and Public Policy, Carnegie Mellon University. Pittsburgh,
PA. 15213.
Direct correspondence to:

Ronald Fernandes
Doctoral Candidate
H. John Heinz III School of Public Policy & Management,
Carnegie Mellon University,
Pittsburgh, PA 15213-3890. U.S.A
Phone: (412) 421 6753:E-mail: ronald@andrew.cmu.edu
We are grateful to the many people working in the Indian software industry who
cooperated in the collection of our data and in allowing us to interview them. The
customary disclaimers on errors and inadequacies apply. This research was supported by
a grant from the Sloan Foundation for research on the software industry in India.
Abstract
Indias success and worldwide reputation in software development and services
has naturally led to questions about the fundamental factors influencing this success. In
this paper, we focus on one very critical factor human capital. We analyze the aspects
related to the supply and demand for skilled human capital-software developers to the
Indian software development industry. We then look at two empirical studies to
understand the ground realities of the challenges facing the industry vis--vis the
acquisition and retention of skilled human capital as it exists right now and prospects for
the future. Our general finding is that a situation of excess demand currently exists for
skilled software development professionals. The large loss of experienced professionals
due to relocation to the U.S. under the H-1 B visa program has exacerbated the problem.
We find that although there are some interesting differences in the types of skills used by
the two major categories of software professionals: engineers and non-engineers with a
certificate from private training institutes, these differences are not very large at the
outset of their job experiences. This indicates that unobserved differences in the
perceived abilities of the individuals rather than educational experience or training could
play a primary role in the skill sets acquired by experienced software professionals.
However, the differences in skills sets acquired with experience do not indicate many
critical differences in the software development capabilities of professionals from these
two categories other than a specialization of skill sets acquired. Given the slow pace at
which engineering education capacity can be expanded, we recommend expanding the
potential pool of recruits beyond its current reliance on engineers and computer science
graduates to include graduates with training from private institutes in order to sustain the
growth rate of the industry.
1. Introduction
The rapid growth of the Indian software industry owes much to lowered barriers
to the international movement of capital, goods and services. Indeed, the Indian software
got its start because of the excess demand for software programmers in the U.S. many
Indian software exporters began by supplying programmers to U.S. firms. Since then the
business models have evolved but the essential source of competitive advantage for the
Indian software industry remains its access to a large source of low cost software
developers (Arora et al 2000).
Despite the large pool of talent available in India, there are concerns that the rapid
growth of demand worldwide and in India itself in particular, will soon lead to a situation
of excess demand. Responding to the demands of the information technology industry in
the U.S., the passage of the U.S. Workplace Competitiveness Act of 1998 permitted an
increase in the supply of skilled software developers to the U.S. economy for a limited
period. A large proportion of this supply is skilled software developers from India. In
India, government policy looking to encourage the continuing spectacular growth of the
industry perceived that future shortfall in the supply of skilled labor was a serious threat.
Recognizing the importance of this fact many Indian policy makers have called
for declaring a state of Educational emergency1 to ensure that the supply of skilled
software developers is increased. Several CEOs of smaller software development firms
and NASSCOM-the professional association representing the views of these firms- have
begun to argue that the shortage of skilled labor is constraining their ability to grow.
Basic Background Report (BR-3) for the National Task Force on Information Technology (IT) and
Software Development (SD) submitted to the Prime Minister of India: 18th March 1999
Larger firms are concerned mainly with a mismatch in the skills of software developers
available, especially for experienced software developers.
Concerns about an impending shortage are not new. For instance, in a study of
the Indian software industry published in 1994, Lakha (1994: 394-396) notes that
although India had 2.5 million scientists and engineers, with 160,000 new ones produced
each year, and despite the many new computer courses introduced by the government in
various institutes, personnel requirements are exceeding the current supply, making it
more difficult to attain the governments software export target. (p 394). Similar
concerns about an actual or impending shortage have been voiced by a number of others,
such as Barr and Tessler (1996).
The basic issue is that the software industry in India has relied heavily upon a
large pool of engineers. However, the growth of the industry and the relatively inelastic
supply of engineering graduates are rapidly exhausting the available pool of engineers.
Therefore, whether the Indian software industry can maintain a high rate of growth in the
future depends critically upon whether it can substitute bright, English speaking
graduates without engineering degrees to make up for the shortfall (relative to growing
demand) in the supply of engineering graduates. The bulk of these new recruits will be
trained in software development and related skills by private training institutes. The
ability to sustain growth also depends on how Indian firms mature and acquire domain
expertise that can offset the diminished labor cost advantage. However, to do so they
must deal with the steady drain of experienced and talented software developers to the
U.S. and other countries.
In this paper, we focus on the overall demand supply situation facing the Indian
software industry. Section 1 discusses how important the labor cost advantage is for
Indian software exporters. Sections 2 & 3 discuss respectively the sources of supply and
demand, concluding with a rough estimate of the projected excess demand if current
hiring patterns hold. Section 4 describes an analysis of a sample of newspaper
advertisements for software developers, substantiating the strong bias in favor of
engineers. Section 5 draws on the analysis of a questionnaire survey administered to 61
software developers to address the differences in the on-the-job skills used by engineers
and non-engineers, and how these differences evolve over time. Section 6 summarizes
our findings and concludes.
1.
The importance of Human Capital for the Software Development
Industry2
The key to the success of Indian software development firms is the supply of
trained, low cost software developers. Estimated wage costs for software developers in
India are about 1/3rd to 1/5th the corresponding U.S. levels for comparable work (INFAC:
Indian Software Market Status Report, 1998). Other estimates suggest that the difference is
smaller, and that once all costs are factored in, the human capital costs of software
development are around of U.S. levels. A variant on the labor cost argument is the
large pool of suitable workers.
In a companion study, we interviewed U.S. firms that had outsourced software
development to India (Arora et al 2000). The most frequently cited reasons for
This section draws heavily upon Arora et al. (2000).
outsourcing have to do with the shortage of software developers in the U.S. - Firms claim
that they simply cannot find enough software developers fast enough. In some cases,
U.S. firms outsource to Indian firms to get access to more specialized engineering talent,
particularly in the area of telecommunications. Other reasons include the option of round
the clock operations and the ability of Indian vendors to assemble "functional" teams of
engineers at a very short notice.
Most of the U.S. managers we interviewed commented on the excellent
programming and coding skills available in India. They also noted that their Indian
vendors were good and willing learners, receptive to new ideas, and flexible in terms of
the software and hardware platforms for which they provide services. Many of the
respondents downplayed the cost difference, insisting that cost was a relatively minor
consideration. This appears to be a response to the prevailing concern in the U.S. about
the possible harm to U.S. engineers from software outsourcing and inflow of foreign
software developers. The apparent lack of concern with cost is inconsistent with the
extensive price competition that prevails in the Indian software industry, and with the
actual behavior of some of the firms that outsource to India.3
In recent years, a new category of software services, called IT-enabled services,
has shown strong growth. This category includes customer interaction services including
call centers and help desks, data entry, medical transcription, travel reservations, content
development and animation and financial transactions for banks and other financial
For instance, one of the telecom firm we interviewed had, in addition to a couple of prominent Indian
firms, a firm largely in the business of bodyshopping among its list of preferred vendors. Our interview
suggested that this was a way of promoting competition and keeping prices down. This firms policy was
to ensure that its business accounted for a substantial (but less than 50%) share of the vendors revenues
and that no single vendor had more than 25% of its business. These policies ensured that it had
considerable bargaining power in the pricing negotiations with its Indian vendors.
institutions. Table 4 (Appendix 1) shows the projections obtained by a NASSCOM

survey that has forecast opportunities for IT-enabled services from India. Currently
around 23,000 software developers are employed in this category, which is anticipated to
show phenomenal growth4 in the next decade. If this anticipated growth occurs, it would
put further pressure on the ability of software development firms to meet their staffing
needs, although some portion of this demand would draw on a different pool of software
developers whose primary skills are related to the use of software rather than software
development5.
Despite the apparently large stock of human capital in India, the National
Association for Software and Service Companies (NASSCOM 1999) claims that by the
year 2000 demand will outstrip supply. The NASSCOM study projects an annual demand
of about 75,000 by the year 2000. In fact, our research leads U.S. to conjecture that,
within segments of skills and experience in software development firms, demand already
far exceeds supply. The growth of the software development industry in India is being
severely constrained due to the shortage in supply of skilled software developers with
According to a recent survey done by NASSCOM (the National Association of Software and Service
Companies), IT-enabled services in India have shown the highest growth (66%) over the previous year
among all the industry segments and are expected to gross a revenue of $ 880 million (Rs 4000 crores) in
2000-01, as against $ 530 million (Rs 2,400 crore) in the previous year. The two most promising segments
this year are customer interaction services including call centers and content development and animation.
The survey indicates that customer interaction services are expected to generate $ 160 million (Rs 750
crores) in this year, compared to $90 million (Rs 400 crore) during the last year. Revenues from content
development and animation are expected to almost double from $ 180 million (Rs 820 crore) in 1999-00 to
$ 355 million (Rs 1,600 crore) in 2000-01. The survey indicates that the following cities: Gurgaon,
Mumbai, Hyderabad, Bangalore, Chennai, Pune, Ahmedabad and Thiruvananthapuram are emerging as
major hubs of IT-enabled services.
5
Medical transcriptions and other forms of data entry for instance simply rely on high school or college
graduates who know typing and language skills (for which they are trained). For instance, to transcribe the
audio taped diagnosis of physicians located in the U.S., Indian firms rely on software developers trained to
understand English spoken with an American accent and who have the ability to type in English.
experience, caused largely due to the loss (immigration) of experienced software

developers with 4-6 years of work experience to the U.S through the H-1B visa route6.
In an earlier study (Arora et al 2000) , we surveyed over a 100 software firms on
various attributes of managing software development firms in India. When asked in
1998-99 to list the top 3 problems they faced, more than half of all firms irrespective of
age, size or market orientations (either export or import) selected manpower shortage and
employee attrition as the most serious problem affecting them.
Overall, software export firms appear to view both manpower shortage and
employee attrition as a more serious problem as compared to firms primarily catering to
the domestic industry. Probably, this is because it is software export firms that have
shown phenomenal growth in the last decade. The Indian software development industry
currently revolves around export with very few firms solely focused on domestic
business. Table 1 shows the number of firms who indicated the seriousness of problems
facing them. Manpower shortage/skilled manpower shortage was considered a more
serious problem than the availability of physical or commercial infrastructure, a reflection
of the criticality of manpower for the software development industry.
Around 30,000 software developers, most of whom are believed to be software developers are believed to
have availed of H1-B visas in 1999- Interview of Mr. Clyde Jones-Chief of Consular Services, U.S
Consulate India and reported in The Economic Times, Hyderabad edition Wednesday 30th June 1999. In
line with what we mention later in the report about the distribution of engineering colleges being heavily
loaded towards states in South India, Chennai, (the only state in the south that has a U.S Consulate)
reported a H1-B issuance of around 20,000 such visas out of the 30,000 that were issued this year from the
three U.S Consulates (Mumbai, Chennai and Calcutta) and the U.S Embassy (New Delhi). The quota of
115,000 visas approved for this year (September 98-September 1999) by the U.S Workplace and
Competitiveness Act 1998 was filled by May 1999.
Table 1: Major problems for Indian software firms

Problem
Export firms Domestic firms
Manpower shortage/skills
57
32
Employee attrition
44
27
Physical Infrastructure
12
12
Commercial infrastructure
24
17
Quality certification
11
Visas
33
NA
Finance/ Capital
20
14
Marketing access
42
17
Lack of domestic computerization
21
Lack of government support
10
11
Tariffs and other barriers
11
Note: The firms were asked to indicate their top three problems.
Source: CMUSoft Dataset -- CMU Survey of Indian Software Industry. N = 104
A strong preference for hiring engineers is a key to the potential excess demand
for labor. Although India graduates over 155,000 engineers a year, there is high demand
for software engineers (or engineers potentially trainable as software engineers), and
wages have been rising at about 20% per annum (NASSCOM 1999). Despite paying
substantially above Indian standards, virtually all firms find it difficult to attract and
retain talented software developers.
Most software development firms recruit either engineers7 or students with
degrees in mathematics or science. Many also have in-house testing and training
Bachelor/Master of Engineering (B.E/M.E), Bachelor/Master of Technology (B.Tech. /M.Tech.) or

Master of Computer Applications (MCA).
programs. All of the firms we interviewed claimed to recruit engineers alone, or

engineers and students with a first degree in mathematics or science.
A potentially serious constraint on the ability to rapidly increase the output of
trained software and computer engineers is the shortage of engineering doctorates being
awarded in India. Recent data show that the PhDs awarded in engineering disciplines
have fallen from their high of 675 in 1987 to 375 in 1995. Concurrently, the number of
engineers with postgraduate training has also risen only slowly, from a little over 12,000
in 1987-89 to a little over 17,000 in 1990-92. Surveys of Indias premier technological
institutions-the Indian Institutes of Technology (IITs) show that a very large fraction of
postgraduates from those institutions enter the Information Technology (IT) sector, in
some cases as many as 90%! Although there are a substantial number of engineering
doctorate holders of Indian origin working in universities and research institutes outside
India who can be called on, a long run solution will require an increase in the
postgraduate research and training infrastructure.
There are number of public sector and industry initiatives to address this
perceived shortage of software developers. The government has recently announced the
establishment of Indian Institutes of Information Technology, along the lines of the wellknown Indian Institutes of Technology. A number of private sector for-profit, training
institutes are also coming on line. Examples of this are institutes set up by Silverline, and
Mahindra-BT, a joint venture between the Mahindra Group and British Telecom. The
latter is an ambitious effort, where the objective is to teach methodologies and techniques
for software development. The partners include the Software Engineering Institute, and
Carnegie Mellon University.
10
Firms are responding to the problem of employee attrition in a number of ways.

One popular way is by providing opportunities for career growth for their employees.
Some firms stress their ability to provide a career path for their employees, wherein they
could move to being managers and would not have to remain programmers, apparently
something valued by Indian software developers. Interestingly enough, few claimed to
pay more than their competitors to attract and retain workers. A number of firms were
actively considering stock options for their employees. Some of the firms expressed the
need to make the organization individual-independent by addressing the loss of
knowledge due to employee turnover. The loss of employees is especially severe among
consultants (programmers and analysts) sent overseas to work onsite in the U.S., and our
interviewees assert that a very substantial fraction quit within two or three years.
Firms do hire workers who have graduated from private training institutes (PTIs)
but without an engineering degree. There however seems to be a perception among the
software industry about the quality of graduates from private training institutes and the
potential negative signals this could send to firms contracting work with them. Despite
the apparent unwillingness of software development firms to hire them, there are a large
number of students trained by private, for-profit, training institutes such as the National
Institute for Information Technology Ltd. (NIIT) and Aptech Ltd., resulting in high
profits and growth for these firms. The (software development) training industry has
grown sharply along with the software industry, with total revenues in 1997-98 estimated
at around $300 million, up from $150 million the year before. Although nearly 90% of
11
revenues are from the individual training sector, corporate training is also growing
rapidly.8
Many of the students trained by private training institutes are apparently
employed in the IT departments and divisions of Indian industry, such as in banking &
financial institutions, and in IT enabled services such as back office operations, Revenue
accounting, data-entry conversion, remote maintenance and support, medical
transcription, insurance claim processing, call centers, database services, and content
development and as educators in training institutes.
We also found evidence that some large and reputable firms had tie-ups for
professional practice with graduates from private training institutes and frequently ended
up recruiting these graduates. The increase in certification courses offered by institutions
such as the Microsoft Certified Software Engineer/Developer (MCSE/D), Certified
Novell Engineer (CNE), IBM Net Professional certification has also attracted a number
of engineers interested in a software development career.
2. Supply
The Indian Software development industry has in the last two decades of its
growth focused primarily on the supply of software services to foreign countries9. A large
chunk of all exports is to the U.S and Europe where English is the language of choice.
Thus, one of the important factors affecting supply of skilled software developers to
Many multinationals have started authorized training centers (ATC) that provide their own certified
courses. These include IBM Global services, Oracle, Microsoft, Adobe, Cadence, Lotus, and Sony.
9
One estimate suggests that India has 16% of the global market in customised software, and that more
than 100 of the Fortune 500 had outsourced to India (Dataquest, 31 July, 96; pp 43-44).
12
software development firms in any country has been the availability of English-speaking
software developers. India currently has the largest English speaking scientific labor
population in the world.1 In particular, the supply of engineering software developers
from what is characterized as the formal system of education in India has played a vital
role in the development and growth of the industry. However, as discussed later in this
paper, a number of software developers with non-engineering backgrounds also enter
software development firms after being trained and certified by private training
institutions.
2.1.
Supply of engineering graduates

The supply of graduates in the engineering fields is governed by policies made by
the All India Council for Technical Education (AICTE). 10 Figures 1 and 2 show the
growth in the number of engineering colleges and student enrollment over the last fifty
years. The number of AICTE approved technical degree institutions in the 1990s has
grown at an average annual rate of 9 % with a jump of around 16% from 1996 to 1997
and 21% from 1997 to 1998.
10
The AICTE Act, 1987 provides for the establishment of an All India Council for Technical Education
(AICTE) to plan and coordinate the development of the technical education system in the country, and
maintain norms and standards for technical education. One important bureau in the council called the
Bureau of Engineering and Technology concerns itself with the accreditation of new programs, approval of
new institutions, sanction of new courses, and sanction of additional intake of the University Grants
programs, post-graduate programs, and MCA/MBA programs.
13
Figure 1: Growth in Number of Engineering (degree and diploma) institutions in India
Source: AICTE web site: http://www.aicte.com/approved_inst.html
Figure 2: Growth in Student Intake for Engineering (degree and diploma) institutions in
India, 1940-1997
The growth in student enrollment in AICTE approved degree institutions in the

1990s has grown at an average rate of 11.5% per annum with a growth of 29% from
1996-1997 and 19% from 1997-1998. Table 2 (Appendix 1) shows the breakup by state,
14
of the number of institutions and student intake into diploma and degree programs in
engineering as well as into the MBA (Master of Business Administration), MCA (Master
of Computer Applications) degree programs, and other post-graduate courses in 1998.
Engineering institutes from southern and southwestern India account for over half of the
engineering graduates (82,000 of 156,000 graduates).
One can also derive a rough estimate of the number of graduates in engineering
and technology available to join the Indian industry in 1998. From a total intake of
156000 individuals, we assume a drop-out rate of 5% i.e., 7800 individuals, a loss after
graduation to other technical post graduate programs of around 20,000 graduates, another
loss to MBA or MCA programs of about 25,000 (which is about 40% of the intake into
these programs). This leaves a balance of approximately 104,000 engineering graduates
available for employment in all sectors of the economy, including the Information
Technology (IT) industry. In addition, we have around 10,000 post-graduates with MCA
degrees most of whom join the IT industry, along with some fraction of the 52,000 MBA
graduates and 20,000 graduates with other post-graduate degrees.
In addition to the annual flow, the software industry has also drawn from the
stock of engineers and technologists. The recession in other sectors of Indian industry in
the 1990s has led to a substantial number of these graduates and post-graduates joining
software development firms. However, we do not have any way of estimating the
number of engineers and managers that have made such a move.
15
2.2. Supply of software developers from private training institutions

Private training institutions train individuals specifically for work in software
development (NASSCOM 1999). For instance NIIT, one of the largest such training
institutions has around 10,000 students (in 1999) in its GNIIT (Graduate NIIT11) program,
a three year program for software developers including 1 year of internship (professional
practice) with a software development firm. Similarly Aptech which claims to be the
worlds largest software training organization in terms of number of students (it has
1200+ franchisees and a presence in 40 countries including the U.S, Europe and
Australia) has an arm, Asset International, that specializes in training software
development software developers. Asset International, the high-end training arm of
Aptech currently trains both software developers with work experience and recent
graduates in certification course such as Microsoft Certified Systems Engineers (MCSE)12
and Microsoft Certified Systems Developer (MCSD), Certified Novell professional
(CNP), Oracle and IBM net commerce software developers. In 1999 around 30% of the
7,200 software developers Asset International trained were engineering graduates. They
11
A play on words because according to current Indian laws, private training institutes are not allowed to
award degrees.
12
Last year around 50,000 software developers obtained the MCSEs certification in the world, of which
around 35-40,000 were certified from India (Interview with the Regional Manager, Asset International).
Certification for most of the software development firms such as Microsoft, Oracle and Lotus is conducted
by Sylvan Prometric, an Australian firm that makes available online the tests required for certification for a
certain fee. The results are collected and transmitted to the various development firms such as IBM, Oracle
for evaluation. Since the cost of obtaining training and certification is quite expensive in the U.S, Indian
software developers often travel to India to renew their certification. These certifications are globally valid
and apart from giving a pass/fail status also give the marks obtained in the test as an indication of the
aptitude of the student. This certification is typically restricted a specific version of the software on which
the student is tested. Whenever a newer version of the product is released the development firm usually
gives a period of 6 months within which the professional needs to upgrade his skills and obtain a recertification of his skills in that software by undergoing another certification exam. Thus, for example, the
initial Oracle certification involves being tested in 5 modules or 5 independent tests and the re-certification
involves giving an additional skills upgrade test.
16
expect to train an additional 5000 students annually in their recently introduced (in 1999)
course in e-commerce.
NASSCOM sources estimated that there were 3800 such training firms in 1998, in
what was then a $300 million market, although together NIIT and Aptech are believed to
have 70% of the software training market. 13 The general perception in the industry
(NASSCOM, 1999; Arora et al, 2000) seems to be that graduates from private training
institutes are not suitable for software development, unless they also have engineering
degrees or have interned with software development firms during the course of their
private training. An earlier study (NASSCOM, 1999) reported that only 2% of all
software developers trained in private training institutes join software development firms.
In our study, however we found instances where graduates from private training
institutes, sometimes through tie-ups with development firms for professional practice14
did end up joining the development industry.
One should also note that private training institutes are also important for helping
existing software developers acquire new skills. As discussed below, many engineers
13
Other major vendors include Software Solutions Integrated, LCC Infotech, Tata Infotech, CMC, Indian
Institute of Hardware Technology, First Computers, Pentafour Communications, Jetking, IIS Infotech,
Boston Education, SQL Star, Datapro and IBM Learning Services.
14
NIIT for instance offers a 3- year GNIIT program that includes 1 year of professional practice
(internship) with a software development firm. Software developers trained with this degree usually end up
being offered a job with the firm on completing the internship. Ramco, a large and one of the first
successful Indian developers of software package and ERP package called Marshall, has recruited software
developers using this source and there are currently around 100 such software developers working with
them. Other firms too appear to recruit such professional. For instance, CITIL, a software development firm
recently used the placement services of Aptech and recruited around 100 Aptech students for various areas
of application. Similarly Hexaware Software, which was initially a part of Aptech but which was
subsequently spun-off as an independent entity directly contacts the placement cell of Aptech and recruits
students who have completed various certificate courses with Aptech. A large number of the graduates
however end up being employed as teachers/trainers in these non-formal institutions or start up private
training institutions of their own. Similarly IIS Infotech, a Delhi based software development firms
(recently taken over by a British group) employs 20-30% graduates from NIIT/Aptech. (Source- Interviews
with project leaders/managers/CEOs of Ramco and Hexaware, the COO of NIIT and the regional manager
of Asset International (a subsidiary of Aptech)
17
also undertake further training in software development on their own from private
training institutes. This points to the rapidly changing nature of the industry and the high
rate of obsolescence of narrow, technical skills, such as knowledge of a specific
programming language.
2.3. Supply of software developers with Government certification

The Department of Electronics through the Department of Electronics
Accreditation of Computer Courses (DOEACC) provides accreditation to private training
institutes in the non-formal sector ranging from foundation level courses to a postgraduate level in the computer field. A number of small private training institutes rely on
this method of accreditation15. This is viewed as one way of establishing their credibility
in the market and is frequently advertised as such.
2.4. Supply of software developers returning from the U.S

In the last two years there appears to be a phenomena where software developers
who have worked in the U.S return to India to either work with firms or get involved in
new start-ups. In interviews conducted in December of 1997, we interviewed several
software developers who returned for personal or professional reasons. We believe that
15
The DOEACC scheme provides accreditation at 4 levels: the O level, which is the foundation level to the
post-graduate C level. The B level is considered equivalent to an MCA degree by the industry. The
National center for Software Technology (NCST), a government society, also conducts tests for
competence in software at various levels. The G level tests software developers with B.E/B.Tech/MCA
degrees and postgraduate degrees in software technology. The C level is for B.Sc, Computer Science
degree holders (BCS) and graduates or equivalents in any discipline who have done a minimum of three
months full-time or six months part-time training and are aspiring for programmer level jobs, and E level
which is an entry level exam open to all graduates to test their aptitude for computer-related training.
18
this flow of software developers back to India who have worked in the U.S. on H-1B16
visas and are now unable to obtain green cards will increase in the future especially due
to the current backlog in issuing green cards17. However there is some concern about the
specific skills such software developers have gained in the U.S. and how relevant these
skills would be in Indian software development firms.
Software developers in India are frequently asked to take on a greater
responsibilities vis--vis their training and educational backgrounds. Thus a professional
with around two years of experience is frequently expected to take on the role of a project
leader, a job which usually involves skills such as performing requirements analysis and
systems analysis and design in addition to the coding, debugging and testing that is the
usual job of an entry level software programmer. Individuals sent to the U.S. on an H1-B
visa are typically not given this level of responsibility and frequently end up performing
fairly narrow and specialized programming activities. On their return to India, these
individuals tend to expect a managerial position for which, according to the managers we
interviewed, they sometimes lack the required experience.
16
The temporary work visa, called H1-B, is a category for non-immigrant foreign workers to work in a
specialized field in the U.S. for up to six years. Almost half of those visa holders are working in the U.S. IT
industry as engineers, computer programmers or medical software developers.
17
Also see Vic Goel: The Green Card Waiting Game, In Silicon India - Technology and Business
Magazine, May 1999 pp.54-56. Another issue is that while H-1B visas are issued to qualified software
developers on which basis they can bring their families to the U.S, green cards have to be issued separately
to individual members of a family.
19
3. Demand
3.1 Changes in demand:
The explosive growth of software development firms in the last decade had led to
an increase in the wages paid to software developers, which on average has increased by
20% annually (in 1998)18. The software development industry is a very human capitalintensive industry and the projected 50% annual growth rates are approximately
comparable to the growth rates in demand for workers.
3.2 Changes in employer preferences

In the beginning, software development firms had a distinct preference for
engineering graduates from the prestigious Indian Institutes of Technology. However,
not only were such graduates difficult to recruit than graduates of lesser known
engineering colleges, they were more likely leave the firm to study or work in the U.S.
Many firms then decided to focus on engineering graduates from regional and state
engineering colleges. Subsequently firms have also begun to consider graduates in
computer science from non-engineering degree colleges (Bachelor of Computer Science
or BCS) and software developers with a post-graduate degree in computer science
(Master of Computer Science or MCS). More recently, smaller firms have also begun to
recruit software developers with any graduate degree but with a certificate or diploma
from private training institutes.
For software developers at the entry level in software firms the general perception
among firms appears to be that 2-3 months of training can provide an individual with the
18
Of basic pay. This did not include any payment of bonus/perks, employee stock options (ESOP) etc.
20
required skills to begin entry-level software development tasks such as programming.

Thus, while any graduate is potentially a suitable candidate for a job in a software
development firm, there still seems to be a distinct preference for engineering graduates,
not just engineering graduates with a specialization in the computer sciences or related
fields but even engineering graduates from unrelated fields such as mechanical, chemical,
or metallurgical engineering. We shall later show that this preference appeared to be
strong for recruiting entry-level software developers but not for experienced software
developers. Our empirical evidence discussed later suggests that this is not what firms
actually prefer but they are constrained by the shortage of software developers trained in
computer science, software development and related fields.
3.3
Migration overseas
A large number of skilled software developers have traditionally gone to the U.S.
every year, a factor that has significantly contributed to the high attrition rates in Indian
software firms and consequently, the shortage of skilled software developers at the
project leader and project manager levels. In the last few years, this number has gone up
even further due to the severe shortage of skilled software developers in the U.S. and
consequent modification in the U.S. Workplace Competitiveness Act of 1998 and an
increase in the H-1 B quota from 65,000 to 115,000 per year. Last year around 20,00019
19
Most of the firms sponsoring these visas were located in the U.S Thus for instance, in 1998, Infosys
Technologies Ltd. (ITL), one of the large software firms based in India had 231 employees working in the
U.S on H-1 visas and another 111 employees on L-1 (intra-company transferee) visas. (Source: Form F-1
of ITL filed with the SEC). Last year 46% of all H-1 visas issued were to software developers from India.
Most of these visas were for skilled software developers based upon our study of the firms that sponsored
these visas-Mastech (11%), Computerpeople (6%), Oracle (5%), Pricewaterhouse Coopers (4%) and
Lucent Technologies (3%) based in the U.S. Tata Consultancy Services based in India accounted for 7% of
all H-1B visas issued. (Source: The Indian Express, Pune, July 14, 1999). Countries with the highest
21
software developers relocated to the U.S. Last year (1999) around 30,000 software
developers20 have been reported to obtain H-1 B visas from India. Most of these software
developers have some experience working in software development firms in India.
Another category of visas used by firms to send software developers to the U.S is the B-1
business visitor visa which permits a short term stay for business purposes in the U.S.
Anecdotal evidence however suggests that it is possible for a B-1 business visa to be
converted to an H-1B visa quite easily and there appear to be no yearly limits or
restrictions on the numbers of such converted visas. Thus, the actual number of software
software developers coming to the U.S from India in 1999 would be much higher than the
estimates given earlier.
3.4 Demand Estimates

NASSCOM estimates the current demand by software development firms to be in
the region of 140,000 software developers for 1999 and 2000. 21 The total stock of
software developers employed was believed to be around 160,000 in 1998 of which 70%
or 112,000 are estimated to be in software development. Based on productivity measures
we can estimate the annual demand by the software industry as follows: In 1998 the total
revenue to the software industry, both development and services was around $4 billion.
From earlier studies (Arora, et. al., 2000), the average annual productivity of software
developers in the industry is estimated at about $0.035 million. Dividing $4 billion by
percentage of total recipients of H-1B visas: India -46% China -10% Canada- 4% Philippines-3%
(Valbrun, 1999)
20
Interview with Mr. Vivek Savant, C-DAC, Interview of Mr. Clyde Jones-Chief of Consular Services,
U.S Consulate India and reported in the The Economic Times, Hyderabad edition Wednesday 30th June
1999.
21
Approximately 56,000 in 1999 and 84,000 in 2000 (assuming a growth rate of 50%).
22
$0.035 million, we get around 115,000 software developers, which also matches the
NASSCOM estimate of 112,000. Assuming a 50% growth in revenues for 1999, this
implies that the industry needs to recruit 57,500 software developers. To this, one must
add the need to replace those lost to the U.S. software industry. In 1998, around 20,000
software developers moved to the U.S., and the number is likely to be larger for 1999.
This implies a total demand of about 75,000 to 85,000 developers.
By comparison, NASSCOMs estimate of the total supply in 1999 is about 70,000 as
shown in Table 3 (Appendix 1). This figure includes around 15,000 graduates in
Computer Science, Electronics and Telecommunications, another 10,000 engineering
graduates with degrees unrelated to software development22, around 10,000 MCA
graduates, 10,000 graduates with B.Sc. (Computer Science) and B.Sc. (Math) degrees. In
addition, there are about 20,000 software developers with diplomas in Computer Science
or Electronics, with certificates in specialized software development areas from NIIT,
Aptech or other private institutes, and around 10,000 software developers with postgraduate degrees such as a M.Sc., an M.Tech. or an MBA. If current hiring practices
continue, this implies an excess demand of 15,000 in 1999.
Shortfall: 15,000
Supply: 70,000
Demand:
75,000 - 85,000
22
H-1B outflow 30,000
We think that the NASSCOM estimates are low as they consider only 41 engineering colleges as
potential sources of engineering graduates for the software industry- India currently has around 662 public
23
Firms have responded to this shortfall so far in three ways. First, they have recruited
more entry-level employees from engineering and science disciplines other than
computer science, electrical and electronics engineering, software engineering and related
disciplines. Second, they have tried to recruit software developers from other sectors with
relevant work experience such as project management, although the quantitative
significance of this move is uncertain in terms of numbers of software developers who
have moved from other industries to the software development industry. Finally, firms
have increasingly turned to software developers trained in software development by
private training institutes.
The projected excess demand, driven largely by an excess demand for engineers,
itself raises the question of why, given the nature of work, engineers are preferred so
strongly over non-engineers? Though the Indian software industry tends to primarily
recruit engineering graduates, their motivations for doing so vary among firms. The bulk
of the work is relatively non-technical and requires mostly logical and methodical work
and a familiarity with software development tools and languages. Data from the firms
surveyed in Arora et al (2000) show that about 17% of the software developers had been
recruited from private training institutes and did not have an engineering degree. Of the
rest, over 70% had an engineering or advanced scientific degree. Few of the firms we
interviewed admitted to hiring any non-engineers, although since the bulk of the
engineering graduates are not trained in software engineering or computer science, a
substantial fraction of the hires also have diplomas from private training institutes.
(state) and private engineering colleges (Table 2, Source: AICTE web site:
http://www.aicte.com/approved_inst.html).
24
Further, some of the smaller and younger firms do hire graduates from these institutions,
as do some domestic market focused firms.
This preference for engineers is even more puzzling since the bulk of the work is
relatively non-technical and requires mostly logical and methodical work and a
familiarity with software development tools and languages. Given the large number of
science and arts graduates, and the widespread availability of private training institutions,
the pool of potential software developers is much larger than merely graduates from
engineering colleges. However, some software firms are reluctant to tap this pool
because of the potential negative signals to their customers.23. As one CEO we
interviewed put it
Take somebody from a good college (any of the top 20 colleges in India), give
him 3 months of orientation and they are ready to take up a programming assignment. I
dont need all these engineers. But I dont want to be branded by my customers as a
guy who hires NIIT graduates. (Emphases added.)
This is a clear instance of a race to the top rather than a race to the bottom.
With only limited market power, Indian software exporters try to distinguish themselves
from the competition by pointing to the quality of their processes and people, and when
possible, their experience. Firms also have quality concerns. Managers we interviewed
believe that an engineering education imparts a set of problem solving skills, methods of
thinking logically and learning tools that help quick adaptation to changes in technology,
domains and tasks. Since Indian firms provide services across a range of platforms and
domains, this is an important asset. The CEO of a small firm developing innovative
products stated:
23
U.S. visa restrictions are another reason why firms prefer engineering graduates.
25
[B] ecause things are changing so fast in this industry, knowledge of a particular
operating system, a particular language, a particular technology is not as important as the
ability to learn and adapt to change.
Software exporters also prefer engineering graduates because they are more likely
to satisfy the U.S. H1-B visa restrictions that require technical credentials. Engineering
graduates are much more likely to have 16 years of education compared to the typical BA
or BS graduate. The preference for engineers may also be due to the high screening
ability of the Indian education system admission to engineering colleges is seen as
indicative of academic ability and aptitude.
The growing presence of private training institutes in cities in India is increasingly
making it possible for software developers to obtain certificates and diplomas from such
institutes. Earlier experiences of the software development industry with such graduates
were not very favorable primarily because of the quality concerns related to the
educational process in such institutes. However, in the last two years the certification of
software development software developers has increased dramatically. Since the testing
and certification is carried out by an independent authority (Sylvan Prometric Inc.) a
Australian based firm that conducts the tests and then contracts with individual reputed
software firms such as Microsoft, Novell and IBM who then verify the results and award
certification. This has led to some level of quality being established for private training
institutes, which is useful for graduates from such institutes to signal their quality. If this
succeeds, it is probable that in the future, graduates from such institutes would be
considered as substitutes to engineers and other graduates from formal institutions.
Private training institutes are also being increasingly being used by graduates from
26
private and state engineering colleges to obtain better jobs offers from software
development firms.
To investigate further, we analyzed two types of data. First, we analyze a dataset
developed from newspaper advertisements in India to recruit software developers. Since
the advertisements could specify the types of skills and formal qualifications, these data
can illuminate the attributes of expressed demand for different types of skills required by
software firms. The second dataset is based on responses to a short questionnaire survey
administered to 61 software developers working in the Indian software industry.
Interviews with CEOs and HR managers of software firms, representatives from private
training institutions, and policy experts complement quantitative evidence on the current
reality of the manpower situation.
4. Analyzing demand of the Indian software industry for different

types of skills using advertisement data
4.1 Research method
Data were collected from advertisements placed by software firms in The Times
of India and The Economic Times, two leading English dailies and published
simultaneously from Mumbai, Bangalore and Delhi. The advertisements were to recruit
software developers, and were from firms from the following industries: Computer
Applications, Computer Networking, Computer Systems, Banking & Insurance,
Consultancy, Telecommunications, Internet and New Media and Business Services and
hosted on the Jobs & Careers website.24 We selected advertisements that specified the job
functions to be Software Development, Consulting or Systems Development. The three
24
http://www.timesjobsandcareers.com/jobsrc2s.html.
27
cities Mumbai, Bangalore and Delhi are large metropolitan cities and have large
concentrations of software development firms. Between 26th April 1999 and 26th June
1999, a period of two months, we found 102 advertisements for software developers in
these seven industries. The coding criterion employed to analyze the data is given in
Table 5 (Appendix 1).
4.2 Results
A large percentage of the advertisements (68%) appeared to be from the computer
systems industry25 followed by the Internet and New Media industry (15%) and
Consultancy industry26 as shown in Table 6. The job specifications described in the
advertisements were predominantly related to software development (82%), followed by
system development (12%) and consultancy (6%).
Table 6: Distribution of Advertisements across Industry
Industry
Frequency
Percent
COMPUTER APPLICATION
COMPUTER NETWORKING
COMPUTER SYSTEMS
69
68
CONSULTANCY
TELECOMMUNICATIONS
INTERNET & NEW MEDIA
15
15
TOTAL
102
Source: Jobs and Careers website of the Times of India and The Economic Times
(1999)
The distribution of advertisements across industries with the job specification as software
development is very similar to the distribution of the entire sample. In view of the possible
25
Some names include: Nucleus Software Exports Ltd, Akiko Sherman Infotech, Datamatics Staffing
Services, Infosys Technologies Limited, Network Programs India Pvt. Ltd and SiCore Systems (P) Ltd,
26
Some names include: PriceWaterHouseCoopers Limited, A.F. Ferguson & Co., Sai Corporate Services
Pvt. Ltd., Text 100 India and ABC Consultants Pvt. Ltd.
28
ambiguity in descriptions of other job specifications and the limited focus of this study,
we limited our subsequent analysis to advertisements by software and IT (information
technology) related firms and where the advertisement was explicitly for software
developers.
Of the 84 advertisements for software developers, only 48, or a little over half of
the advertisement explicitly specified formal educational qualifications. Table 7 shows
that 36 (75%) of these 48 job listings also specified an alternative qualification. The first
qualification, specified by 83% of the advertisements was for software developers with a
graduate technical degree (B.E or B.Tech), followed by 8% for a post-graduate technical
degree (M.E or M.Tech.) and another 8% for any graduate degree. The second
qualification (Table 8) specified by 42% of the advertisements was an MCA degree, 25%
specified an M.E or M.Tech., and 17% asked for advanced diplomas in a specialized field
of software development such as Multimedia.
Table 7: First Qualification specified in Job listing Advertisements for Software

Development
First Qualification
Frequency
Percent
B.E\B.TECH
40
83
M.E\M.TECH
OTHER GRAD
TOTAL
48
(1999)
29
Table 8: Second Qualification specified in Job listing Advertisements for Software

Development
Second Qualification Frequency Percent
DIPLOMA
17
B.E\B.TECH
11
M.E\M.TECH
25
M.C.A
15
42
B.SC
TOTAL
36
(1999)
Of the 48 advertisements that specified qualifications, 24 (50%) also mentioned the

fields of specialization. As Table 9 shows, 58 % required a specialization in Computer
Sciences & Engineering; another 17% required a specialization in Electronics and
Communications, and another 17% in Multimedia.
Table 9: Specialization fields for advertisements that stipulate qualifications
Specialization
Frequency
Percent
MECHANICAL
COMP.SC\ENGG
14
58
ELECTRICAL
ELEC & COMM
17
MULTIMEDIA
17
TOTAL
24
(1999)
The average work experience required was specified was 5.23 years indicating a
need for slightly more experienced software developers than the current average
experience as determined by other sources (NASSCOM 1998).
30
These data show a clear preference among employers for engineering graduates: a
B.E or B.Tech degree was strongly preferred as the first qualification of choice followed
by an MCA as the second qualification of choice. Interestingly, most employers that
required an engineering degree also specified a specialization in computer sciences &
engineering, or electronics or communications, and most of the remaining in Multimedia,
thus indicating that the preference for engineering graduates was not for graduates with
specialization in any field but in specific fields clearly related to the software
development field. Additionally the specialization in Multimedia required by employers
is something that at present is mostly provided by non-formal training institutions and
thus it is possible that this category of employees recruited mostly included graduates in
engineering who also had diplomas or certificates from private training institutes.
One interesting finding is that 52% of advertisements did not specifically mention
any education qualification requirement but instead just described the project and the skill
sets required. One possible explanation is that only employers who wanted to recruit
experienced software developers, who might not have the specified qualifications and
thus might be hesitant to respond to the advertisement, may have advertised. To test if
this was indeed the case, we correlated the experience variable with a dummy variable
that indicated whether the advertisement specified a qualification requirement. We
computed a Pearsons coefficient r = 0.08 which indicated a very low correlation between
these two variables. Thus, the failure to indicate specific educational qualifications is not
simply because of a desire to recruit more experienced software developers.
From the above we can infer that either a) employers are desperately in need of
people hence they do not wish to restrict their options by specifying a particular
31
qualification, or b) employers believe that irrespective of formal qualifications, they can

train software developers whom they recruit to perform the tasks to which they are
assigned, or, c) that neither formal nor informal qualifications matter, the skill sets or the
ability of software developers as observed during the recruitment process suffice to
identify suitable employees. All in all this points to a fairly chaotic and unsystematic
method of recruitment indicative of the young age of the industry and the turbulent times
it is passing through.
5. Data from personal questionnaires

To evaluate the match of skills provided by the educational system with the
requirement of the industry, we then interviewed software developers that had worked on
at least a couple of projects. We focused on project experience rather than experience as a
function of duration (such as six monthly or annually) as this would better help U.S.
understand the skills and domain knowledge that software developers gained during their
work experience. It was also cognitively less trying for a professional to detail the skills
he or she used in a project because this is the usual information they would have on their
resume or curriculum vitae. We collected information on the first to the seventh projects
(beginning with the first software development related project) on which the respondent
worked. Our choice of number of projects was guided by a couple of considerations.
First, since the rate of obsolescence of technical knowledge is high, the relevance of the
educational experience of software professionals would be the highest in the first few
years (six or seven projects) of joining the industry. Subsequently corporate training and
experience would affect the technical knowledge of a professional and affect our
32
evaluation of the match between the needs of the software industry and the supply of
human capital. This could also happen if education was not the critical variable but onthe job training after employment was what mattered. Secondly, we needed to keep the
questionnaire short. We selected software developers employed in either software export
firms or domestic software firms at different locations in India. We wanted to explore the
differences in skill and capability between engineers and non-engineers as software
developers.
5.1 Research method

We interviewed 61 software developers in Pune, Madras, Bangalore, and
Hyderabad, which are major software centers in India.27 Face to face or telephonic
interviews enabled us to collect information on the specific skills and abilities, and their
links to the educational training of the individual. We selected from major categories of
software developers (B.E, B.Tech., M.E, M.Tech., MCA, B.Sc, M.Sc., certificate from
private training institute) as given in Table 4. At least 5 individuals in each category were
interviewed using a structured questionnaire format (Appendix 2). The selection of firms
was done randomly in these four cities. However, within a firm, we specifically (over
sampled) software developers who had graduated from a private training institute. In
addition, we interviewed professionals who directly managed software developers (such
as project managers & project leaders) or were indirectly associated with their selection,
recruitment and career growth (such as human resource managers) to ascertain their
views of the factors affecting demand and supply of software developers using an openended interview format. We also interviewed industry experts and senior management
33
software developers from the two largest non-formal training institutions in India viz.
NIIT and APTECH. These interviews were carried out face-to-face or telephonically over
the summer of 1999 in India.
In the structured questionnaire (See Appendix 2) administered to software
developers, the specific questions related to the following aspects:
i)
Demographics information such as age and sex
ii)
Educational information such as educational qualifications after high school

(both formal and non formal), specializations chosen and the year of graduation.
iii)
Project information (for each of the first 7 projects in the software developers
career), including the description, size and duration of the project, a description
of the tasks and programming languages, technologies, and skills used, and from
where those skills were attained (e.g., college, on the job exposure, on the job
training program, private training or certification program and or own effort)
5.2 Preliminary statistics for the sample

Of the 61 software developers in the sample, 11 (18.03%) were female while 50
were male (81.97), which corresponds to NASSCOM figures for 1998 as well. The
sample mean age was 26.5 years while the median age was 25 years. NASSCOM figures
for 1998 put the median age for software developers in the industry at 26.6 years. The
median duration of projects in the sample was 8 months while the median work
experience was 3 years. NASSCOM estimated the median working experience for IT
workers at around 5.9 years, thus indicating that our sample had software developers who
27
The interviews were carried out by the first listed author.
34
were somewhat less experienced than the national average. The median size of the
average project was around 60 man months. These 61 software developers had worked
for 111 firms and on 230 software development projects by the time of their interview.
The list of firms in which these software developers worked for one or more projects
during their first seven projects is given in Appendix 3.
The breakdown of initial studies undertaken by software developers in the sample
after high school is shown below in Table 10. Most software developers have a B.E,
B.Tech. or a B.Sc. (Computer Science) [56%] followed by a B.Sc. (Non Computer
Science) [26%]. The highest numbers of specializations were computer science or
engineering (31%) followed by the other category (29%). Recall however that we over
sampled non-engineers and therefore these figures are not representative of the industry
as a whole.
Table 10: Distribution of First Educational Qualification of Sample Software

Developers
First Educational Qualification
Frequency
Percent
DIPLOMA
B.E/B.TECH/B.SC (COMP. SCIENCE)
34
56
B.SC (NON-COMP. SCIENCE)
16
26
OTHER GRAD
10
TOTAL
61
Source: Carnegie Mellon Software Developers Skills Questionnaire (1999)
On average, the sample software developers have 41.3 months (3 years) of

undergraduate education and 26.5 months of formal postgraduate education (for those
that did have graduate education). Finally, about one-third of the sample also had 19
35
months of non-formal post-graduate education. Table 11 shows the average duration for
formal undergraduate studies, formal and non-formal graduate studies and mostly nonformal studies undertaken in private training institutions.
Table 11: Duration of Formal and Informal Education, Sample Software
Developers
Duration of Education n
Mean
s.d
Min Max
FORMAL-UNDERGAD
59
41.3
6.31
36
60
FORMAL GRAD
39
26.5
11.45
48
NON-FORMAL POST-GRAD
18
19.1
14.9
48
Table 11 highlights the extensive education software developers pursue, with more than
half of the sample having formal graduate degrees (MCA most frequent) or non-formal
certificates from private training institutes and a quarter of the sample also having some
additional form of non-formal post-graduate education. The subsequent graduate studies
undertaken by around 70% of software developers in the sample was more varied than
the initial undergraduate studies as shown in Table 12
Table 12: Post-graduate educational qualification for sample software developers
Second Educational Qualification
Frequency
Percent
ME/M.TECH
MCA
14
MBA
18
41
OTHER FORMAL GRADUATE/POST-GRADUATE
14
OTHER
M.SC (COMP. SCIENCE)
CERTIFICATE FROM PRIVATE TRAINING
INSTITUTIONS
43
TOTAL
36
Our figures (Table 13) for software developers with non-formal certificates from
private training institutes as their second educational qualification are on the higher side
than would be representative of software development firms in India. This over sampling
was done on purpose to enable us, as part of our research design, to compare project
experience related differences between software developers who had formal engineering
or Computer science degrees vis--vis those who had other formal graduate degrees from
private training institutes.
Table 13: Post-graduate educational qualification for sample software developersNon-Engineering/Non-Comp. Science graduates vis--vis Engineering/ Comp.
Science graduates.
Second Educational
Qualification
Non-Engineering/NonComp. Science graduates
Engineering/ Comp.
Science graduates
Frequency
Percent
Frequency
Percent
MCA
26
MBA
16
15
63
16
42
OTHER FORMAL
GRADUATE/POST-GRADUATE
25
OTHER
TOTAL
24
CERTIFICATE FROM PRIVATE
TRAINING INSTITUTIONS
ME/M.TECH/
SCIENCE)
M.SC (COMP.
19
Of the 34 software developers in our sample with engineering or computer science

degrees, 19 (56%) undertook further education as compared to software developers in our
sample without these degrees of whom 24 out of the 37 or around 65% undertook further
education. We also observe that only 16 % of those with an engineering or computer
science degree went for a diploma or certificate from a private training institute with the
majority of them opting for an engineering or computer science related graduate degree,
37
compared with around 60% for those without an engineering degree. The average
duration of studies undertaken on a full time basis for all software developers during the
attainment of their second educational qualification was 26.5 months (slightly more than
2 years) with software developers with engineering degrees spending around 28 months
on average as compared to 25 months for software developers without engineering
degrees.
An additional 17 software developers, a larger proportion of whom were software
developers without engineering degrees also had a third educational qualification with the
distribution shown in Table 14. The largest number (59%) was of individuals obtaining
training and certificates in software development from private training institutions with
an average duration of around 18 months.
Table 14: Distribution of third educational qualifications among software
developers.
Third Educational Qualification
Frequency
Percent
MCA
MBA
12
10
59
OTHER FORMAL GRADUATE/POST-GRADUATE
18
OTHER
TOTAL
17
CERTIFICATE FROM PRIVATE TRAINING
INSTITUTIONS
Thus, we observe that a) professionals with engineering or related degrees (B.ScComputer Science) appear to be better off in terms of the extent of educational
background needed. b) A large proportion of professionals in our sample appeared to
have spent a substantial amount of time training in non-formal training institutions, and c)
38
professionals without formal engineering degrees appear to be very willing to undertake

an extensive amount of education prior to joining software development firms.
In general, our descriptive statistics reveal that a) software developers with
engineering or related degrees (B.Sc-Computer Science) appear to be better off in terms
of the extent of educational background needed. b) a large proportion of software
developers in our sample appeared to have spent a substantial amount of time training in
non-formal training institutions, and c) software developers without formal engineering
degrees appear to be very likely to have undertaken an extensive amount of education
prior to joining software development firms.
5.3 Detailed Analysis

In order to examine the differences between engineering graduates and those who had
certificates from private training institutions we created two categories of software
developers in our dataset. The first included all software developers who received an
engineering graduate or post-graduate degree at one or more of the three educational
levels, with no exposure to private training institutions at any level. The second included
all software developers who had a diploma or certification from a private training
institution in their second or third educational levels and with a non-engineering degree
in their first educational level. The first category of software developers with engineering
degrees (n=28), were then compared with the second category of software developers
with degrees from private training institutions (n=24). Software developers with both
engineering degrees and certificates from private training institutes (n=9) were not
included in the analysis as they had a substantial exposure both to engineering and formal
39
training institutions making it impossible for us to distinguish analytically whether their

current skill set and employment characteristics were due to the one or the other.
A comparison (Table 15, columns 3 and 4) of the average duration of projects
worked on (in months) showed no sizeable differences with the two categories working
an average of 8.1 months on each project. However there was a large and significant
(0.05 level of significance) difference in the average size of projects worked on (in manmonths) with software developers with an engineering degree working on projects with
an average size of 220 man-months while those with diplomas or certificates from private
training institutes working on projects with an average size of 87 man months.
Table 15: Difference between PTI grads and Engineers on Job characteristics
(1)
(2)
(3)
PROFESSIONAL
AVERAGE
PROJECT
CATEGORY
DURATION
(MONTHS)
PTI GRADS
24
8.1
ENGINEERS
28
8.1
(4)
AVERAGE
PROJECT SIZE
(MAN-MONTHS)
87
220**
(5)
(6)
(7)
AVERAGE
AVERAGE
AVERAGE
PROJECTS
JOBS
EXPERIENCE
W ORKED
W ORKED
(YEARS)
(# OF
PROJECTS)
(# OF JOBS)
3.5
1.5
2.9
3.9
2.1**
3.7
** Significant at 0.05 level.
Table 15 also shows (columns 5, 6 and 7) that there was no significant difference in the
average number of projects worked or in the average experience at the time of the survey
between graduates from private training institutes and engineering college graduates in
our sample. However there was a significant difference (at the 0.05 level) in the average
number of jobs worked on during the course of these projects, with engineering graduates
having worked on a higher average number of jobs as compared to graduates from private
40
training institutes. This indicates lower job mobility for graduates from private training
institutes as compared to software developers with an engineering degree in our sample.
In general, the median length of time spent in one job by software developers in our
sample was 2 years. Alternatively, Table 16 shows the distribution in our sample by
number of firms worked for the whole sample, PTI graduates and Engineers. For 66% of
PTI graduates their current job was the only one they had worked in whereas this was
true for around 43% of engineering graduates.
Table 16: Difference between proportion of PTI grads and Engineers on # of Jobs
worked
JOBS W ORKED
ALL
PTI GRADS
ENGINEERS
52
66
43
23
21
25
15
13
18
Software developers Skills Analysis

We then coded the languages and skills used by both these groups of software developers
for each of the projects they worked. These were then recoded into skill categories based
upon similarities between them as shown in Table 17:
41
Table 17: Languages and Skills comprising a skill category

CODE
SKILL CATEGORY
LANGUAGES/SKILLS
SCRIPTING LANGUAGES
ACTIVEX, CGI, HTML, JAVASCRIPT, MS ASP, VB SCRIPT,

XML, LOTUS SCRIPT
OBJECT ORIENTED LANGUAGES
C++, JAVA, OOPS, VC++
PROCEDURAL LANGUAGES
BASIC, C, COBOL, PASCAL, PERL, LEX/YACC

CASE TOOL, CCCA TOOL, GUI TOOL, ER-IN TOOL, SEEC TOOL, CT TOOL
TOOLS
DATABASE SKILLS
PLATFORM SKILLS
MTS, RTOS
NT DEVICE DRIVER, NT INTERNALS, UNIX
SYSTEM LEVEL CODING SKILLS
INTERNALS
ARCHITECTURE KNOWLEDGE
COM, CORBA, MAPI
9
10
OPERATING SYSTEMS
APPLICATIONS
CICS, DOS, JCL, MVS, UNIX, PL1

ADOBE PHOTOSHOP
11
DEVELOPER TOOLS
POWERBUILDER, VISUAL BASIC, DEVELOPER2000
ORACLE, FOXPRO, GUPTA SQL, IMS, INFORMIX, INGRESS, MS

ACCESS, FOXPRO, UDB, PL/SQL
Our subsequent analysis looked at project-wise differences in proportions of skill

categories used by both categories of software developers, the change in the use of these
skills within a category and the effect of time (as measured by changes from one project
to the next) on changes in the proportions of use of these skills. Table 18 shows how the
proportion of use of the top five skill categories changed between PTI graduates and
engineers across their first six projects. (These top five skill categories accounted for
more than 80% of the skills used by both categories of software developers across all
projects). For instance, the topmost cell reading from the left is 0.05 or 5% indicating that
in the sample, the percentage of software developers from private training institutes using
scripting language skills in their first project (P1) was 5 percentage points more than the
percentage of engineering software developers using scripting language skills in their
first project, although the difference was not significant.
42
Table 18: Difference between PTI graduates and engineers in proportions of skill
categories used across projects
SKILL CATEGORY
P1
P2
P3
P4
P5
P6
SCRIPTING LANG.
0.05
-0.03
0.06
-0.21*
-0.27***
-0.31**
OBJECT ORIENTED LANG.
0.06
0.20**
0.02
0.14
0.33**
-0.08
PROCEDURAL LANG.
-0.09
-0.14*
-0.15**
-0.03
0.01
0.32
DATABASE LANG.
-0.04
0.07
-0.01
0.14
0.17
DATABASE SKILLS
0.01
-0.08**
-0.09**
-0.10*
*** Significant at the 0.01 level; ** Significant at the 0.05 level; * Significant at the 0.10 level
At the first project level, there appear to be no significant differences in the skills used by
engineering versus PTI graduates. By the second project, some significant differences
emerge with PTI graduates using more of object-oriented language skills and less of
procedural language and database skills as compared to engineering graduates. With the
passage of time and a gaining in project experience for all software developers, we see
that the scripting language and the database skill category shows a marked decline across
projects with more engineering graduates ending up using this skill by the end of the
sixth project as compared to software developers from private training institutes.
To examine the changes in these skills within the software developers categories
we were examining we then averaged the proportions of each of the top five skill
categories across the first and second project and the sixth and seventh project and then
computed the changes in proportions between these two averages (Table 19).
Table 19: Difference between PTI graduates and engineers in proportions (average)
of skill categories used between the First/Second and Sixth/Seventh Project.
Skill Category
PTI
Engineers
Scripting Lang.
-0.15***
0.02
Object Oriented Lang.
-0.29***
0.06
Procedural Lang.
0.08
Database Lang.
0.30**
Database Skills
-0.02
-0.21***
0.08
-0.02
43
For example, the topmost cell on the left indicates that the average proportion of software
developers from private training institutes who used scripting language skills in their first
and second projects decreased by 15 percentage points (significant at the 0.01 level)
between their first/second projects and their sixth/seventh projects. Overall, differences
emerge in the changes in skill categories with the use of scripting and object oriented
languages decreasing but the use of database language skills increasing for PTI graduates.
On the other hand, engineers do not exhibit any significant differences in the changes in
skills categories between the first two and the last two projects in our sample except in
the use of procedural language skills (Basic, C, COBOL, Pascal, PERL, LEX/YACC) where
the change is negative and significant.
To examine in more detail the effect of shifting from one project to the next on
the change in skill categories used by both categories of software developers we then
regressed the proportion of each skill categories on project worked (Table 20). We see
that for PTI graduates, the proportion of use of scripting language skills declines by 3.6%
over time (measured in projects) and of object-oriented skills declines by 4.5% while
database language skills increase by 5.9%. For engineers on the other hand a similar
change results in a decrease of 3.8% in the use of procedural language skills and a 2.2%
increase in database language skills. Database skills for both categories of software
developers show almost no change across projects.
44
Table 20: Regression of proportion of use of skill category on project

Skill Category
PTI
Engineers
2
Coeff
Coeff
Scripting Lang.
-3.6**
0.58
0.2
0.00
Object Oriented Lang.
-4.5*
0.45
1.0
0.04
1.8
0.13
-3.8***
0.89
Database Lang.
5.9***
0.79
2.2*
0.52
Database Skills
-0.5
0.38
0.0
0.00
Procedural Lang.
In general, these findings point to some interesting differences in skills.

However, whether these differences reflect training or inherent (or perceived) differences
in ability between the two groups is an open question. One explanation of this difference
could be the differences in average project sizes worked on by the two types of software
developers as being indicative of the complexity of projects and requiring a higher order
of skills. Another possible explanation is that Scripting and Object Oriented languages
need a greater problem-solving ability and logical thinking capabilities for which
engineers are better equipped to handle, or at least, are perceived to better.
6. Conclusions
Our interviews with industry experts and industry heads and the analyses
presented above points to the mismatch between supply and demand for labor in the
Indian software industry. The substantial preference for engineers combined with rapid
growth in demand and the relatively inelastic supply of engineers are the basic cause of
the mismatch. A substantial attrition of experienced software developers to the U.S.,
under the H1-B visa scheme, which has dramatically increased over the past few years
exacerbates the situation. As a CEO of the software development firm bluntly put it, We
45
dont have an infinitely large supply of software developers here. What I know is that
whatever the supply, there will still be a demand for it. The recession in India has also
contributed to this by making the software industry an attractive recruiter even for
graduates with a non-related degree. However there is no certainty that these favorable
conditions will continue in the future.
The mutual attraction between engineers of unrelated disciplines and the software
industry has serious implications for other sectors of the Indian economy. The AICTE in
a recent report talks about the unstoppable exodus of students from the parent
engineering disciplines to the IT sector, resulting in an acute shortage in these
disciplines.'' Engineering institutes are now believed to produce around 7,000 M Techs a
year, a mere one-third of their capacity. The number of doctorates in engineering has also
fallen drastically raising serious concerns about the falling research standards in
engineering and technology and a potential shortage of qualified faculty in engineering
colleges.
We find that although there are some interesting differences in the types of skills
used by engineers and non-engineers, these differences are not very large at the outset. If
the differences were primarily due to training, rather than unobserved differences in the
perceived abilities of the individuals, one might have expected greater differences in the
skill sets used during the earlier projects. Further research is needed to probe whether the
preference for engineers is merely because admission to engineering school is a useful
screening device or whether engineering training develops the problem solving and other
skills useful for software development. Even so, it is clear that given the slow pace at
which engineering education capacity can be expanded, sustaining the growth rate of the
46
Indian software industry will require expanding the potential pool of recruits beyond
engineers and computer science graduates. India has a large pool of college graduates, a
reasonable fraction of which can be trained to develop software. Rising wages for
software developers are already providing incentives for employers to tap sources other
than engineers in order to close the gap between demand and supply. Until now, the gap
is the smallest at the entry level. There is a much greater shortage at the experienced
project leader and project manager level thus leading to a) software developers with
inadequate experience being compelled to handle these tasks with job quality and
employee motivational consequences, and b) firms being unable to undertake new tasks
and move up the value chain, thus inhibiting growth, productivity and profitability. This
shortage is largely because a large number of experienced software developers (4-6 years
of experience) who move to the U.S are potential managerial material for firms. The
drastic loss of these software developers (a conservative estimate of which is around
30,000 software developers28 in 1999, of whom many could be presumed to be
experienced given the stringent criteria established for issuing an H1-B visa) has the
potential for creating a serious problem for software development firms even in the short
run. The opportunity costs of the loss of these software developers who usually leave
after working around 3-4 years in the firm include projects forsaken due to a shortage of
qualified project leaders or managers, the disincentive of software firms to impart
training to software developers and poorer quality software. Given the nearly guaranteed
loss of experienced software developers whose work for Indian firms is viewed as a
stepping stone to a job in the U.S, firms may decide in the future not to invest in the
28
A higher estimate of this number in 1999 is around 71,000 software developers (Arya, 1999) although
current estimates are that nearly 50% of last years quota of 115,000 H1-B visas went to Indians.
47
identification and training of engineers with managerial potential. Any change in the
current situation (recession) in other sectors of the Indian economy could lead to a
shortage of graduates for the software development industry at the entry level as well.
48
References
All India Council for Technical Education website: http://www.aicte.com
Annual Report 1996-97: DOEACC Society (A society under Department of Electronics,
Government of India)
Aptech Limiteds People Annual Report for 1998
Arora, Ashish, V.S. Arunachalam, Jai Asundi, Ronald Fernandes (2000)The Indian
Software Industry Heinz School Working Paper, Carnegie Mellon University,
Pittsburgh.
Arya, Gaurav Cap on Foreign Computer Pros in U.S. Is Hit --- Enrollments at Two
Schools In India Are Unlikely To Suffer Due to Demand, Wall Street Journal, Eastern
edition, New York, May 24, 1999
Basic Background Report (BR-3) for the National Task Force on IT and SD submitted to
the Prime Minister of India: 18th March 1999. http://it-taskforce.nic.in/it-taskforce/
Bischoff, Laura A. The United States Finds Itself Tapping a Worldwide Workforce,
[CITY Edition], Dayton Daily News, Dayton, July 11, 1999
Branigin, William A Push for More Special Work Visas; Tech Employers Say Higher
Limit on Foreigners Isn't High Enough Washington Post, June 25, 1999
CareerPath.com Continues Popular Section on Immigration Issues; Ongoing
Chats Underscore Trend towards Online Immigration, PR Newswire, New York
July 27, 1999.
INFAC, Indian Software Market Status, Report, January 1998 pp. 15, Bombay.
Lakha, Salim, 1994, The new international division of labor and the Indian computer
software industry, Modern Asian Studies, 28, 2, pp 381-408.
Looking for a Certificate IDC Directions 2000 Indias 12th Annual IT Industry Briefing
Session July 1999.
Marquand, Robert India's tropics: Where the Engineers Are, Christian Science
Monitor, September 14, 1999
Minutes of the meeting of the National Task Force on Information Technology and
Software Development. Venue: Office of the Chief Minister, Andhra Pradesh
Secretariat, Hyderabad. 21 November 1998
49
Minutes of the meeting of the National Task Force on Information Technology and
Software Development. Venue: Raj Bhavan, Calcutta, 3rd August 1998
Minutes of the Meeting of the National Task Force. Venue: Conference Room of
Sahyadri, Government of Maharashtra, Mumbai 18 June 1998.
National Association of Software and Service Companies (NASSCOM), The Indian Software
Industry in India: A Strategic Review January 1999, New Delhi, India.
Radhika Sachdev Steps to beef up engineering studies urged The Times of India,
November 4, 1999 http://www.timesofindia.com/today/04indi9.htm
Rosenberg, Ronald High-Tech SOS is answered from afar [SUNRISE Edition], The
Oregonian, Portland, Oregon. March 14, 1999
Transcript of interview with Vivek Savant, Program coordinator, C-DAC and member of
the National Task Force on Information Technology and Software Development-HRD
working group.
Transcript of interview with Mr. Chetan Shetty, President, Ruksun Software, Pune
Transcript of interview with Mr. George Paul, Regional Manager, Asset International (a
division of Aptech)
Transcript of interview with Mr. Srikant Inamdar, VP (Software Development group),
NIIT, Delhi
Transcript of interview with Mr. Rajendran, Chief Operations Officer, NIIT, Delhi
The Software Industry in India: A Strategic Review 1999. NASSCOM, New Delhi 110
021. India
Shyamanuja Das Looking for a Certificate? Cyber India Online July 1999
On the Training Mall Computers Today May 1995 pp. 56-59
The Quality Quandary Computers Today May 1995 pp. 61-62
Valbrun, Marjorie Renewed Bids for Visas for High-Tech Workers Reflect the
Political Influence of Silicon Valley, Wall Street Journal, September 15, 1999
Valbrun, Marjorie Fraud Concerns Grow Over Employment-Visa Program, Wall Street
Journal, Eastern edition, New York, July 12, 1999
Vic Goel: The Green Card Waiting Game, In Silicon India - Technology and Business
Magazine, May 1999 pp.54-56
50
Washington Probes Fraud in Issuance of Employment Visas, The Indian Express, Pune
Wednesday, July 14th, 1999
Which Course 95? Computers Today May 1995 pp. 67-71
Wipro Launches University Degree Verification in The Economic Times, Hyderabad
edition Wednesday 30th June 1999.
51
Appendix 1
Table 2
Engineering and Managerial Institutions Approved by the AICTE along with
Student intake in 1998.
ZONE & STATE

Central
Madhya Pradesh
Eastern
Arunachal Pradesh
Assam
Andaman & Nicobar
Manipur
Meghalaya
Mizoram
Nagaland
Orissa
Sikkim
Tripura
West Bengal
Northern
Bihar
Uttar Pradesh
North-western
Chandigarh
Haryana
Jammu & Kashmir
New Delhi
Punjab
Rajasthan
Himachal Pradesh
Southern
Andhra Pradesh
Pondicherry
Tamil Nadu
South-western
Karnataka
Kerala
Western
Gujarat
Maharashtra
Goa
Daman & Diu
Engg. &
Tech
(Degree)
Instns.
Engg. &
Tech
(Diploma)
Intake
Instns.
Intake
30
5675
55
7730
1
3
210
660
150
20
3795
1
19
120
3492
1
9
2
3
2
2
1
26
2
1
43
180
1255
170
215
190
260
90
4144
270
125
5475
12
51
2385
9030
29
107
3
27
8
6
19
11
2
530
4095
1060
1490
3810
2129
350
88
2
129
MBA
Instns. Intake
Other
MCA
PG
Instns. Intake Instns. Intake
25
2570
13
390
12
1024
330
60
1
7
54
134
30
25
1128
15
480
487
14
2180
30
10
1377
3824
14345
14
82
1670
7150
1
19
30
580
8
19
804
2217
4
29
12
26
36
24
4
640
4370
2870
4917
6755
2783
695
1
15
5
30
10
20
1
150
1125
240
5600
640
1260
60
120
7
4
400
211
1
2
1
30
60
60
8
6
6
1090
483
614
20285
540
32160
89
4
207
14945
622
43634
108
1
105
5019
60
7565
101
3
106
3050
90
3760
17
4
38
1414
100
2985
70
19
24752
4860
188
50
33025
7880
43
10
2487
510
24
3
770
120
33
5
2021
759
20
118
2
4850
28985
330
220
12230
120
2
11
60
330
10
40
2
1035
2362
85
155743
7400
31215
1100
90
201214
20
115
2
662
38
168
8
1
1171
652
52344
309
10020
242
19656
52
Table 3: Software developers Potentially Available for Employment in Software

Development firms
Degree
Grad
Grad
Grad
Grad
Grad
Undergrad
Undergrad
Technical
Technical
Doctorate (PhDs)
Master of Technology (M.Tech)
Master of Science (M.Sc)
Bachelor/Master of Computer Applications (BCA/MCA)
Post Graduate Diploma
Bachelor of Technology (B. Tech)
Bachelor of Science (B.Sc)
Diploma
ITI/Certificate
Total
Source: The Software Industry in India: A Strategic Review 1999 NASSCOM
Number of
Software
developers
(1996)
95
2,050
2,250
5,050
3,000
14,060
2,800
16,200
15,000
60,505
Number of
Software
developers
(1998)
95
2,130
2,800
7,700
5,000
16,160
3,200
16,700
14,000
67,785
Table 4: Current and Projected Demand for Software developers in IT-enabled

Services
1998
IT-enabled services
Employed
2008 (Projections)
Rs.
Crore
Can be
employed
Rs.
Crore
Back Office operations/ Revenue accounting/data-entry

conversion
9,700
420
2,60,000
19,000
Remote maintenance and support
1,600
65
1,80,000
13,500
Medical Transcription/Insurance claim processing
3,800
140
1,60,000
11,000
Call Centers
1,400
40
1,00,000
6,000
Database services
1,000
45
1,00,000
6,500
Content development
5,500
270
3,00,000
25,000
Total
23,000
980
11,00,000
81,000
Source: The Software Industry in India: A Strategic Review 1999-NASSCOM
53
Table 5
Coding scheme for Advertisement data

INDUSTRY:
1. COMP APPLICATION
2. COMP NETWORKING
3. COMP SYSTEMS
4. BANKING\INSURANCE
5. CONSULTANCY
6. TELECOMMUNICATIONS
7. INTERNET \NEW MEDIA
8. BUSINESS SERVICES
FUNCTIONS:
1. SOFTWARE DEVELOP
2. CONSULTANCY
3. SYS MGMNT
QUALIFICATIONS:
1. DIPLOMA
2. B.E\B.TECH
3. M.E\M.TECH
4. M.C.A
5. B.SC
6. M.SC
7. PH.D
8. OTHER GRAD
9. NIIT
10.APTECH
11.OTHER NON-FORMAL
12.OTHER P.G
13.M.B.A
14.C.A.
15.LAW
SPECIALIZATION:
1. MECHANICAL
2. COMP SC\ENGG
3. ELECTRICAL
4. ELEC & COMM
5. ECO
6. ARTS
7. COMP MGNT
8. MATHS
9. C.A.
10.OTHER
11.SOFTWARE ENGG
12.P.G
13.GRAD
14.P.G. ENGG
15.DIPLOMA
54
Appendix 2
Carnegie Mellon Software Developers Skills

Questionnaire 1999
Name: _________________________________ 2. Age:_____
3. Educational Qualifications:
Institution
Graduation Year
Degree Awarded
4. Work Experience after graduation:
First Project:
Name of Firm:
Project Title:
Project Duration:
months
Project Size:
Duration you worked on Project:

man months
months
Rs./$:
Job Title:
Job Description:
Prog. Languages used:

Skills Used:
(please indicate whether skills
were imparted through:
1: College education
2: Training program
3: On the job exposure
4. Non-formal education
5. Own Efforts
6. Any Other)
Code
Skill
55
Appendix 3
List of firms covered by the survey and # of Software developers who worked in
these firms on various projects in their career.
Firm Name
ACCEL
AERONAUTICAL DEVELOPMENT
AGENCY
AISHWARIYA FREIGHT
CONNECTION
APTECH
ASSET INTERNATIONAL(APTECH)
BEML
BFL SOFTWARE
BITS COMPUTER EDUCATION
BSL SOFTWARE
BIRLA CONSULTANCY
C-DOT
CITIL
CMC
CITATION COMPUTER
CONSULTANTS
CLASSIC TOURS AND TRAVELS
COMPUTER POWER
CYBERMEDIA SOFTWARE
CYBERSOURCE
DSS
DATACOM COMPUTERS
DATACON
DIGICON INDIA
DR. REDDYS LABS
FUNSKOOL SOFTWARE
DEVELOPMENT CENTRE
FUTURA
GCMMF
HCL-HP
HARDCORE (DIV. OF APTECH)
HYDERABAD
IBM
IBM GLOBAL
ICM COMPUTER CONSULTANTS
IDL CHEMICALS
ITSPARK
INDOTRONIKS INTERNATIONAL
INFOMATICS
INFOSYS
INSTITUTE OF PUBLIC
ENTERPRISES
INTELLIGROUP
INTERGRAPHICS
JAYRAM CYBERNETICS
LINK SOFTWARE
LUCENT
MAARS SOFTWARE
MARTIAN COMPUTERS
MASTEK
MICROAIDS
NIIT
NTS COMPUTER CONSULTANTS
OMC COMPUTERS
ORCHID SOFT SYSTEMS
PROGRESSIVE SOFTWARE
RAMCO SYSTEMS
S/W CONSULTING
Frequency
Percent
1
2
0.43
0.87
0.43
1
1
1
3
1
2
1
2
8
2
1
0.43
0.43
0.43
1.3
0.43
0.87
0.43
0.87
3.48
0.87
0.43
1
7
1
1
2
1
1
3
1
1
1
2
4
1
1
0.43
3.04
0.43
0.43
0.87
0.43
0.43
1.3
0.43
0.43
0.43
0.87
1.74
0.43
0.43
1
14
2
1
1
4
1
7
6
0.43
6.09
0.87
0.43
0.43
1.74
0.43
3.04
2.61
13
9
1
1
2
1
2
1
1
1
2
1
1
1
27
4
5.65
3.91
0.43
0.43
0.87
0.43
0.87
0.43
0.43
0.43
0.87
0.43
0.43
0.43
11.74
1.74
56
Firm Name
SSI
SATYAM COMPUTERS
SATYAM INFOWAY
SHERLOCK INFOTECH
SMART SOFTWARE
SONATA SOFTWARE
SYNERGY SOFTWARE
SERVICES
SYNMECH ENTERPRISES
SYSTECH SOLUTIONS
SYSTEM LOGIC
TISL
TVS BRAKES INDIA
TATA INFOTECH
THE HINDU
UUNET INDIA LTD
VALUE SOFTWARE
WIPRO SYSTEMS
Frequency
Percent
1
9
1
10
1
2
1
0.43
3.91
0.43
4.35
0.43
0.87
0.43
1
1
3
1
1
21
1
1
2
12
0.43
0.43
1.3
0.43
0.43
9.13
0.43
0.43
0.87
5.22
TOTAL PROJECTS
230
TOTAL FIRMS
111

Supply and Demand 2001

Transféré par

Informations du document

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

Supply and Demand 2001

Transféré par

Droits d'auteur :

Formats disponibles

Supply and Demand for Software Developers in India

Direct correspondence to:

This section draws heavily upon Arora et al. (2000).

institutions. Table 4 (Appendix 1) shows the projections obtained by a NASSCOM

experience, caused largely due to the loss (immigration) of experienced software

Table 1: Major problems for Indian software firms

Export firms Domestic firms

Lack of domestic computerization

Lack of government support

Tariffs and other barriers

Bachelor/Master of Engineering (B.E/M.E), Bachelor/Master of Technology (B.Tech. /M.Tech.) or

programs. All of the firms we interviewed claimed to recruit engineers alone, or

Firms are responding to the problem of employee attrition in a number of ways.

Supply of engineering graduates

Figure 1: Growth in Number of Engineering (degree and diploma) institutions in India

Source: AICTE web site: http://www.aicte.com/approved_inst.html

Source: AICTE web site: http://www.aicte.com/approved_inst.html

The growth in student enrollment in AICTE approved degree institutions in the

2.2. Supply of software developers from private training institutions

2.3. Supply of software developers with Government certification

2.4. Supply of software developers returning from the U.S

3.2 Changes in employer preferences

required skills to begin entry-level software development tasks such as programming.

3.4 Demand Estimates

H-1B outflow 30,000

4. Analyzing demand of the Indian software industry for different

INTERNET & NEW MEDIA

Table 7: First Qualification specified in Job listing Advertisements for Software

Table 8: Second Qualification specified in Job listing Advertisements for Software

Of the 48 advertisements that specified qualifications, 24 (50%) also mentioned the

ELEC & COMM

qualification, or b) employers believe that irrespective of formal qualifications, they can

5. Data from personal questionnaires

5.1 Research method

Demographics information such as age and sex

Educational information such as educational qualifications after high school

5.2 Preliminary statistics for the sample

The interviews were carried out by the first listed author.

Table 10: Distribution of First Educational Qualification of Sample Software

B.E/B.TECH/B.SC (COMP. SCIENCE)

B.SC (NON-COMP. SCIENCE)

Source: Carnegie Mellon Software Developers Skills Questionnaire (1999)

On average, the sample software developers have 41.3 months (3 years) of

Source: Carnegie Mellon Software Developers Skills Questionnaire (1999)

OTHER FORMAL GRADUATE/POST-GRADUATE

M.SC (COMP. SCIENCE)

CERTIFICATE FROM PRIVATE TRAINING

Source: Carnegie Mellon Software Developers Skills Questionnaire (1999)

Non-Engineering/NonComp. Science graduates

CERTIFICATE FROM PRIVATE

Source: Carnegie Mellon Software Developers Skills Questionnaire (1999)

Of the 34 software developers in our sample with engineering or computer science

OTHER FORMAL GRADUATE/POST-GRADUATE

CERTIFICATE FROM PRIVATE TRAINING

professionals without formal engineering degrees appear to be very willing to undertake

5.3 Detailed Analysis

training institutions making it impossible for us to distinguish analytically whether their

Source: Carnegie Mellon Software Developers Skills Questionnaire (1999)

** Significant at 0.05 level.

Source: Carnegie Mellon Software Developers Skills Questionnaire (1999)

Software developers Skills Analysis

Table 17: Languages and Skills comprising a skill category

ACTIVEX, CGI, HTML, JAVASCRIPT, MS ASP, VB SCRIPT,

OBJECT ORIENTED LANGUAGES